AOAC SPIFAN Stakeholder Program Meeting Book (March 17, 2021)

VII. Precision A. All samples selected for precision studies will be analyzed in duplicate on each of 6 days using multiple analysts and instruments as practical for the different days. Fresh reagents and working standards will be used each day. Reports will include information of number of analysts, instruments, etc. B. Precision data using a CRM or reference material should be included for all candidate methods. The number of matrices may vary between analytes. C. Estimate within-day (repeatability), between-day, and overall precision (intermediate precision; estimation of method uncertainty) for each sample type. Estimates pooled across the sample types may also be useful, if appropriate. Accuracy (Trueness) A. Analysis of CRM or Reference Material 1. A minimum of 9 independent replicates of CRM should be tested on each of 3 days and compared to certified or reference values provided. 2. Where a CRM is not available, a suitable reference material may be substituted. 3. A t -test or other statistical test should be used to evaluate method bias. B. Spike Recovery 1. Recovery will be determined from an appropriate sampling of a range of infant formula and adult nutritional matrices (use of the recognized SPIFAN kit is recommended where appropriate, otherwise the Study Directors may agree on the samples to be included). 2. Each selected matrix will be spiked at 2 levels. For unfortified matrices, use spike levels of 50, 100 and 150% of the typical fortified target. For fortified matrices, use fortification levels of 0%, +50% and +100% of the fortified range. 3. Spiked and unspiked samples will be analyzed in duplicate on each of 3 days. 4. The overall mean of unspiked samples will be used for calculating individual recoveries. [1] No specific criterion in SMPR; recommend calibration errors to be <15%. Along the whole range. Higher calibration error at LLOQ can be accepted ( see FDA 2018, Bioanalytical Method Validation, guidance for industry). [2] Independently prepared standards, if feasible. [3] Concentration of analyte in Matrix Blank/Technical Reagent Blank to be <10% of the estimated LOQ

VIII.